I post a summary of an article from the JMIR this week examining the relative strength of Google Scholar vs. PubMed for medical searches. What struck me on the article (here) was the authors contention that GS provided more free citations. Well today I had the chance to try GS out and was gobsmacked that many many of the citations that came back were behind a pay wall. So I wrote to the lead author. He quickly got back to me very quickly with the following:
In our study, we found that on average only ~25% of the relevant articles (understanding this was a strict definition of relevance – usually RCTs) per search were available for free from Google Scholar and ~16% from PubMed. Unfortunately free access is more an artifact of the journal that published the article than the search engine. However, Google Scholar almost always provides a link to the publisher’s site, while PubMed is hit or miss. Unlike PubMed, Google Scholar also links to non-Journal websites where a full-text article may be posted (may not always be legal).
One trick I use is to maximize the chance of finding the full text of an article is to always click on the “All X Versions” (X representing a number) link at the bottom of the citation beside “Related Articles” in Google Scholar and trying all the links. Additionally, sometimes searching for an article’s title surrounded by quotes in Google proper and limiting to pdf documents can help.
I hope this helps any of you using GS.